结合HHT方法在描述信号瞬时特性和信号处理方面的独特优势，提出了一种基于HHT的泄流结构损伤实时监测诊断方法。首先，利用信号能量谱构造能量分布向量ηi，进而定义结构异常指标和预警指数，最后通过计算相邻时间段内的异常指标，实现结构损伤的在线监测。应用文中方法对一导墙结构进行数值仿真试验，结果表明，文章中定义的异常指标能够较准确地判断结构的损伤，同时具有一定的抗噪能力。

In the long-term service period, many factors may lead to serious damages and deficiencies to the structure, such as complicated working conditions, inadequate investigation and design, poor management, aging of structure, etc. Therefore, it's very important to study the online monitoring method for flood discharge. According to the special advantages of HHT in instantaneous characteristics analysis and signal processing, an online monitoring method for flood discharge structure damage based on HHT is put forward in this paper. Firstly, after obtaining the HHT energy spectrum of measured responses, the energy distribution vector, defined as ηi , can be calculated. Then, because of the η i differences between undamaged and damaged situations, we determine the structure anomaly indexes and the warning threshold. Finally, having been calculating the anomaly indexes of adjacent time periods, the damage online monitoring can be realized. The results from the numerical simulation test of a guide wall demonstrate that the anomaly indexes mentioned above can accurately identify the structure damage and have anti-noise ability to some degree. However, the further research for estimating the damage location and degree need to be developed in the next period.

练继建（1965— ） ， 男，福建建瓯市人，天津大学教授、博士生导师，主要从事水利水电工程安全与优化运行、海上新能源开发利用研究